Data Modeling

Abstract Any organization, whether a business, a government agency, a religion, a family, or any functional human unit, is a jungle of transactions, and these transactions are 99% logical. The vast majority of these transactions are messages sent from senders to receivers. The relationship between these two roles is crucial. A sender does not really know if a message it sends has any value to a receiver. From the standpoint of a receiver, each message it receives falls into one of two categories. If the message resolves some uncertainty in the receiver, or, if it adds to the receiver's knowledge or stimulates action, it is an information message. If it does none of these things, it is a noise message . Information messages, are positive forces in an organization, whereas noise messages are negative forces. The ratio of information messages received to all messages sent is the measure of the information quality of an organization. As this ratio approaches 1, the efficiency and effectiveness of the organization become optimum. This logic commutes the idea of information from a passive concept in organizational theory, to an active force for organizational growth and survival. It is entirely presumptuous to assume that a sender can know for sure whether a message it sends will be an information message or a noise message to a receiver. At best, the sender can only work to increase the probability that its messages will be information messages and not noise messages. And this is where management must focus its attention. A primary aspiration of organizational management must be to increase information quality in the organization by concentrating on increasing the probability that messages sent from its senders to its receivers will be information messages and not noise messages. The question then becomes: “How do we increase the probability of information messages?” The answer, as we shall see, lies in an information engineering technology called Data Modeling. This article covers three aspects of data modeling. First it will cover why we data‐model. Then it will present a sample data modeling technique, and examine how data models are built. Finally, it will discuss nine different applications of data modeling.